vendor/llvm/llvm-trunk-r306956

7 files changed, 99 insertions, 59 deletions

diff --git a/lib/Transforms/Scalar/ConstantHoisting.cpp b/lib/Transforms/Scalar/ConstantHoisting.cpp
index c3810366bf22..a49c9b68c97d 100644
--- a/lib/Transforms/Scalar/ConstantHoisting.cpp
+++ b/lib/Transforms/Scalar/ConstantHoisting.cpp
@@ -38,6 +38,7 @@
 #include "llvm/ADT/SmallVector.h"
 #include "llvm/ADT/Statistic.h"
 #include "llvm/IR/Constants.h"
+#include "llvm/IR/GetElementPtrTypeIterator.h"
 #include "llvm/IR/IntrinsicInst.h"
 #include "llvm/Pass.h"
 #include "llvm/Support/Debug.h"
@@ -340,6 +341,49 @@ void ConstantHoistingPass::collectConstantCandidates(
   }
 }
 
+
+/// \brief Check the operand for instruction Inst at index Idx.
+void ConstantHoistingPass::collectConstantCandidates(
+    ConstCandMapType &ConstCandMap, Instruction *Inst, unsigned Idx) {
+  Value *Opnd = Inst->getOperand(Idx);
+
+  // Visit constant integers.
+  if (auto ConstInt = dyn_cast<ConstantInt>(Opnd)) {
+    collectConstantCandidates(ConstCandMap, Inst, Idx, ConstInt);
+    return;
+  }
+
+  // Visit cast instructions that have constant integers.
+  if (auto CastInst = dyn_cast<Instruction>(Opnd)) {
+    // Only visit cast instructions, which have been skipped. All other
+    // instructions should have already been visited.
+    if (!CastInst->isCast())
+      return;
+
+    if (auto *ConstInt = dyn_cast<ConstantInt>(CastInst->getOperand(0))) {
+      // Pretend the constant is directly used by the instruction and ignore
+      // the cast instruction.
+      collectConstantCandidates(ConstCandMap, Inst, Idx, ConstInt);
+      return;
+    }
+  }
+
+  // Visit constant expressions that have constant integers.
+  if (auto ConstExpr = dyn_cast<ConstantExpr>(Opnd)) {
+    // Only visit constant cast expressions.
+    if (!ConstExpr->isCast())
+      return;
+
+    if (auto ConstInt = dyn_cast<ConstantInt>(ConstExpr->getOperand(0))) {
+      // Pretend the constant is directly used by the instruction and ignore
+      // the constant expression.
+      collectConstantCandidates(ConstCandMap, Inst, Idx, ConstInt);
+      return;
+    }
+  }
+}
+
+
 /// \brief Scan the instruction for expensive integer constants and record them
 /// in the constant candidate vector.
 void ConstantHoistingPass::collectConstantCandidates(
@@ -365,44 +409,25 @@ void ConstantHoistingPass::collectConstantCandidates(
   if (AI && AI->isStaticAlloca())
     return;
 
-  // Scan all operands.
-  for (unsigned Idx = 0, E = Inst->getNumOperands(); Idx != E; ++Idx) {
-    Value *Opnd = Inst->getOperand(Idx);
-
-    // Visit constant integers.
-    if (auto ConstInt = dyn_cast<ConstantInt>(Opnd)) {
-      collectConstantCandidates(ConstCandMap, Inst, Idx, ConstInt);
-      continue;
-    }
-
-    // Visit cast instructions that have constant integers.
-    if (auto CastInst = dyn_cast<Instruction>(Opnd)) {
-      // Only visit cast instructions, which have been skipped. All other
-      // instructions should have already been visited.
-      if (!CastInst->isCast())
-        continue;
-
-      if (auto *ConstInt = dyn_cast<ConstantInt>(CastInst->getOperand(0))) {
-        // Pretend the constant is directly used by the instruction and ignore
-        // the cast instruction.
-        collectConstantCandidates(ConstCandMap, Inst, Idx, ConstInt);
-        continue;
+  // Constants in GEPs that index into a struct type should not be hoisted.
+  if (isa<GetElementPtrInst>(Inst)) {
+    gep_type_iterator GTI = gep_type_begin(Inst);
+
+    // Collect constant for first operand.
+    collectConstantCandidates(ConstCandMap, Inst, 0);
+    // Scan rest operands.
+    for (unsigned Idx = 1, E = Inst->getNumOperands(); Idx != E; ++Idx, ++GTI) {
+      // Only collect constants that index into a non struct type.
+      if (!GTI.isStruct()) {
+        collectConstantCandidates(ConstCandMap, Inst, Idx);
       }
     }
+    return;
+  }
 
-    // Visit constant expressions that have constant integers.
-    if (auto ConstExpr = dyn_cast<ConstantExpr>(Opnd)) {
-      // Only visit constant cast expressions.
-      if (!ConstExpr->isCast())
-        continue;
-
-      if (auto ConstInt = dyn_cast<ConstantInt>(ConstExpr->getOperand(0))) {
-        // Pretend the constant is directly used by the instruction and ignore
-        // the constant expression.
-        collectConstantCandidates(ConstCandMap, Inst, Idx, ConstInt);
-        continue;
-      }
-    }
+  // Scan all operands.
+  for (unsigned Idx = 0, E = Inst->getNumOperands(); Idx != E; ++Idx) {
+    collectConstantCandidates(ConstCandMap, Inst, Idx);
   } // end of for all operands
 }
 
diff --git a/lib/Transforms/Scalar/InductiveRangeCheckElimination.cpp b/lib/Transforms/Scalar/InductiveRangeCheckElimination.cpp
index 2f96c3064b86..a40c22c3fce9 100644
--- a/lib/Transforms/Scalar/InductiveRangeCheckElimination.cpp
+++ b/lib/Transforms/Scalar/InductiveRangeCheckElimination.cpp
@@ -917,7 +917,6 @@ LoopConstrainer::calculateSubRanges() const {
   // I think we can be more aggressive here and make this nuw / nsw if the
   // addition that feeds into the icmp for the latch's terminating branch is nuw
   // / nsw.  In any case, a wrapping 2's complement addition is safe.
-  ConstantInt *One = ConstantInt::get(Ty, 1);
   const SCEV *Start = SE.getSCEV(MainLoopStructure.IndVarStart);
   const SCEV *End = SE.getSCEV(MainLoopStructure.LoopExitAt);
 
@@ -948,8 +947,9 @@ LoopConstrainer::calculateSubRanges() const {
     //    will be an empty range.  Returning an empty range is always safe.
     //
 
-    Smallest = SE.getAddExpr(End, SE.getSCEV(One));
-    Greatest = SE.getAddExpr(Start, SE.getSCEV(One));
+    const SCEV *One = SE.getOne(Ty);
+    Smallest = SE.getAddExpr(End, One);
+    Greatest = SE.getAddExpr(Start, One);
   }
 
   auto Clamp = [this, Smallest, Greatest](const SCEV *S) {
diff --git a/lib/Transforms/Scalar/LoopUnrollPass.cpp b/lib/Transforms/Scalar/LoopUnrollPass.cpp
index 62aa6ee48069..530a68424d5c 100644
--- a/lib/Transforms/Scalar/LoopUnrollPass.cpp
+++ b/lib/Transforms/Scalar/LoopUnrollPass.cpp
@@ -131,7 +131,7 @@ static const unsigned NoThreshold = UINT_MAX;
 /// Gather the various unrolling parameters based on the defaults, compiler
 /// flags, TTI overrides and user specified parameters.
 static TargetTransformInfo::UnrollingPreferences gatherUnrollingPreferences(
-    Loop *L, const TargetTransformInfo &TTI, int OptLevel,
+    Loop *L, ScalarEvolution &SE, const TargetTransformInfo &TTI, int OptLevel,
     Optional<unsigned> UserThreshold, Optional<unsigned> UserCount,
     Optional<bool> UserAllowPartial, Optional<bool> UserRuntime,
     Optional<bool> UserUpperBound) {
@@ -158,7 +158,7 @@ static TargetTransformInfo::UnrollingPreferences gatherUnrollingPreferences(
   UP.AllowPeeling = true;
 
   // Override with any target specific settings
-  TTI.getUnrollingPreferences(L, UP);
+  TTI.getUnrollingPreferences(L, SE, UP);
 
   // Apply size attributes
   if (L->getHeader()->getParent()->optForSize()) {
@@ -699,7 +699,7 @@ static uint64_t getUnrolledLoopSize(
 // Calculates unroll count and writes it to UP.Count.
 static bool computeUnrollCount(
     Loop *L, const TargetTransformInfo &TTI, DominatorTree &DT, LoopInfo *LI,
-    ScalarEvolution *SE, OptimizationRemarkEmitter *ORE, unsigned &TripCount,
+    ScalarEvolution &SE, OptimizationRemarkEmitter *ORE, unsigned &TripCount,
     unsigned MaxTripCount, unsigned &TripMultiple, unsigned LoopSize,
     TargetTransformInfo::UnrollingPreferences &UP, bool &UseUpperBound) {
   // Check for explicit Count.
@@ -770,7 +770,7 @@ static bool computeUnrollCount(
       // helps to remove a significant number of instructions.
       // To check that, run additional analysis on the loop.
       if (Optional<EstimatedUnrollCost> Cost = analyzeLoopUnrollCost(
-              L, FullUnrollTripCount, DT, *SE, TTI,
+              L, FullUnrollTripCount, DT, SE, TTI,
               UP.Threshold * UP.MaxPercentThresholdBoost / 100)) {
         unsigned Boost =
             getFullUnrollBoostingFactor(*Cost, UP.MaxPercentThresholdBoost);
@@ -836,6 +836,8 @@ static bool computeUnrollCount(
     } else {
       UP.Count = TripCount;
     }
+    if (UP.Count > UP.MaxCount)
+      UP.Count = UP.MaxCount;
     if ((PragmaFullUnroll || PragmaEnableUnroll) && TripCount &&
         UP.Count != TripCount)
       ORE->emit(
@@ -926,7 +928,7 @@ static bool computeUnrollCount(
 }
 
 static bool tryToUnrollLoop(Loop *L, DominatorTree &DT, LoopInfo *LI,
-                            ScalarEvolution *SE, const TargetTransformInfo &TTI,
+                            ScalarEvolution &SE, const TargetTransformInfo &TTI,
                             AssumptionCache &AC, OptimizationRemarkEmitter &ORE,
                             bool PreserveLCSSA, int OptLevel,
                             Optional<unsigned> ProvidedCount,
@@ -948,8 +950,8 @@ static bool tryToUnrollLoop(Loop *L, DominatorTree &DT, LoopInfo *LI,
   bool NotDuplicatable;
   bool Convergent;
   TargetTransformInfo::UnrollingPreferences UP = gatherUnrollingPreferences(
-      L, TTI, OptLevel, ProvidedThreshold, ProvidedCount, ProvidedAllowPartial,
-      ProvidedRuntime, ProvidedUpperBound);
+      L, SE, TTI, OptLevel, ProvidedThreshold, ProvidedCount,
+      ProvidedAllowPartial, ProvidedRuntime, ProvidedUpperBound);
   // Exit early if unrolling is disabled.
   if (UP.Threshold == 0 && (!UP.Partial || UP.PartialThreshold == 0))
     return false;
@@ -977,8 +979,8 @@ static bool tryToUnrollLoop(Loop *L, DominatorTree &DT, LoopInfo *LI,
   if (!ExitingBlock || !L->isLoopExiting(ExitingBlock))
     ExitingBlock = L->getExitingBlock();
   if (ExitingBlock) {
-    TripCount = SE->getSmallConstantTripCount(L, ExitingBlock);
-    TripMultiple = SE->getSmallConstantTripMultiple(L, ExitingBlock);
+    TripCount = SE.getSmallConstantTripCount(L, ExitingBlock);
+    TripMultiple = SE.getSmallConstantTripMultiple(L, ExitingBlock);
   }
 
   // If the loop contains a convergent operation, the prelude we'd add
@@ -1000,8 +1002,8 @@ static bool tryToUnrollLoop(Loop *L, DominatorTree &DT, LoopInfo *LI,
   // count.
   bool MaxOrZero = false;
   if (!TripCount) {
-    MaxTripCount = SE->getSmallConstantMaxTripCount(L);
-    MaxOrZero = SE->isBackedgeTakenCountMaxOrZero(L);
+    MaxTripCount = SE.getSmallConstantMaxTripCount(L);
+    MaxOrZero = SE.isBackedgeTakenCountMaxOrZero(L);
     // We can unroll by the upper bound amount if it's generally allowed or if
     // we know that the loop is executed either the upper bound or zero times.
     // (MaxOrZero unrolling keeps only the first loop test, so the number of
@@ -1030,7 +1032,7 @@ static bool tryToUnrollLoop(Loop *L, DominatorTree &DT, LoopInfo *LI,
   // Unroll the loop.
   if (!UnrollLoop(L, UP.Count, TripCount, UP.Force, UP.Runtime,
                   UP.AllowExpensiveTripCount, UseUpperBound, MaxOrZero,
-                  TripMultiple, UP.PeelCount, LI, SE, &DT, &AC, &ORE,
+                  TripMultiple, UP.PeelCount, LI, &SE, &DT, &AC, &ORE,
                   PreserveLCSSA))
     return false;
 
@@ -1073,7 +1075,7 @@ public:
 
     auto &DT = getAnalysis<DominatorTreeWrapperPass>().getDomTree();
     LoopInfo *LI = &getAnalysis<LoopInfoWrapperPass>().getLoopInfo();
-    ScalarEvolution *SE = &getAnalysis<ScalarEvolutionWrapperPass>().getSE();
+    ScalarEvolution &SE = getAnalysis<ScalarEvolutionWrapperPass>().getSE();
     const TargetTransformInfo &TTI =
         getAnalysis<TargetTransformInfoWrapperPass>().getTTI(F);
     auto &AC = getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F);
@@ -1157,7 +1159,7 @@ PreservedAnalyses LoopUnrollPass::run(Loop &L, LoopAnalysisManager &AM,
   if (!AllowPartialUnrolling)
     AllowPartialParam = RuntimeParam = UpperBoundParam = false;
   bool Changed = tryToUnrollLoop(
-      &L, AR.DT, &AR.LI, &AR.SE, AR.TTI, AR.AC, *ORE,
+      &L, AR.DT, &AR.LI, AR.SE, AR.TTI, AR.AC, *ORE,
       /*PreserveLCSSA*/ true, OptLevel, /*Count*/ None,
       /*Threshold*/ None, AllowPartialParam, RuntimeParam, UpperBoundParam);
   if (!Changed)
diff --git a/lib/Transforms/Scalar/NewGVN.cpp b/lib/Transforms/Scalar/NewGVN.cpp
index 7a7624f77542..9cf01c6582b5 100644
--- a/lib/Transforms/Scalar/NewGVN.cpp
+++ b/lib/Transforms/Scalar/NewGVN.cpp
@@ -2423,8 +2423,7 @@ void NewGVN::addPhiOfOps(PHINode *Op, BasicBlock *BB,
   AllTempInstructions.insert(Op);
   PHIOfOpsPHIs[BB].push_back(Op);
   TempToBlock[Op] = BB;
-  if (ExistingValue)
-    RealToTemp[ExistingValue] = Op;
+  RealToTemp[ExistingValue] = Op;
 }
 
 static bool okayForPHIOfOps(const Instruction *I) {
diff --git a/lib/Transforms/Scalar/Reassociate.cpp b/lib/Transforms/Scalar/Reassociate.cpp
index 6da551bd7efd..cdba0062953f 100644
--- a/lib/Transforms/Scalar/Reassociate.cpp
+++ b/lib/Transforms/Scalar/Reassociate.cpp
@@ -1894,6 +1894,8 @@ void ReassociatePass::EraseInst(Instruction *I) {
         Op = Op->user_back();
       RedoInsts.insert(Op);
     }
+
+  MadeChange = true;
 }
 
 // Canonicalize expressions of the following form:
diff --git a/lib/Transforms/Scalar/RewriteStatepointsForGC.cpp b/lib/Transforms/Scalar/RewriteStatepointsForGC.cpp
index a52739bb76f7..a73e9aec0617 100644
--- a/lib/Transforms/Scalar/RewriteStatepointsForGC.cpp
+++ b/lib/Transforms/Scalar/RewriteStatepointsForGC.cpp
@@ -1954,7 +1954,7 @@ static void rematerializeLiveValues(CallSite CS,
       // to identify the newly generated AlternateRootPhi (.base version of phi)
       // and RootOfChain (the original phi node itself) are the same, so that we
       // can rematerialize the gep and casts. This is a workaround for the
-      // deficieny in the findBasePointer algorithm.
+      // deficiency in the findBasePointer algorithm.
       if (!AreEquivalentPhiNodes(*OrigRootPhi, *AlternateRootPhi))
         continue;
       // Now that the phi nodes are proved to be the same, assert that
diff --git a/lib/Transforms/Scalar/SROA.cpp b/lib/Transforms/Scalar/SROA.cpp
index 80fbbeb6829b..4729f4ef5956 100644
--- a/lib/Transforms/Scalar/SROA.cpp
+++ b/lib/Transforms/Scalar/SROA.cpp
@@ -2402,9 +2402,20 @@ private:
       if (LI.isVolatile())
         NewLI->setAtomic(LI.getOrdering(), LI.getSynchScope());
 
+      // Any !nonnull metadata or !range metadata on the old load is also valid
+      // on the new load. This is even true in some cases even when the loads
+      // are different types, for example by mapping !nonnull metadata to
+      // !range metadata by modeling the null pointer constant converted to the
+      // integer type.
+      // FIXME: Add support for range metadata here. Currently the utilities
+      // for this don't propagate range metadata in trivial cases from one
+      // integer load to another, don't handle non-addrspace-0 null pointers
+      // correctly, and don't have any support for mapping ranges as the
+      // integer type becomes winder or narrower.
+      if (MDNode *N = LI.getMetadata(LLVMContext::MD_nonnull))
+        copyNonnullMetadata(LI, N, *NewLI);
+
       // Try to preserve nonnull metadata
-      if (TargetTy->isPointerTy())
-        NewLI->copyMetadata(LI, LLVMContext::MD_nonnull);
       V = NewLI;
 
       // If this is an integer load past the end of the slice (which means the
@@ -3580,10 +3591,11 @@ bool SROA::presplitLoadsAndStores(AllocaInst &AI, AllocaSlices &AS) {
     int Idx = 0, Size = Offsets.Splits.size();
     for (;;) {
       auto *PartTy = Type::getIntNTy(Ty->getContext(), PartSize * 8);
-      auto *PartPtrTy = PartTy->getPointerTo(LI->getPointerAddressSpace());
+      auto AS = LI->getPointerAddressSpace();
+      auto *PartPtrTy = PartTy->getPointerTo(AS);
       LoadInst *PLoad = IRB.CreateAlignedLoad(
           getAdjustedPtr(IRB, DL, BasePtr,
-                         APInt(DL.getPointerSizeInBits(), PartOffset),
+                         APInt(DL.getPointerSizeInBits(AS), PartOffset),
                          PartPtrTy, BasePtr->getName() + "."),
           getAdjustedAlignment(LI, PartOffset, DL), /*IsVolatile*/ false,
           LI->getName());